For gas that never gets much below 145 psig, most of the commercial products do an OK job. They all have quirks (e.g., several versions of PipeFlow had a built in bias that forced flow to go from low numbered nodes to higher numbered nodes and if you messed up on your node numbering you got results that did not match measured values, another wouldn't allow you to input a compressor suction, just discharge and available horsepower). Any model you purchase needs to be verified on an existing line prior to using it to predict a new network or an extension to a network.
For low pressures I use MNET from the now defunct SSI. I probably have the last functioning security device in the world for that software. It is so old that it won't run on a 64 bit operating system so I keep a 16 bit machine around for models. I've tested 40 current-generation programs at low pressures and have not found one that can handle my test cases (i.e., I calibrate the model to match a set of measured parameters in a pipe network, then input another set of measured parameters that happened to reverse the flow in one of the trunks, none of them will reverse the flow in that line and consequently the average wellhead pressure is more than 25% from the actual volume-weighted average). Stoner is the current state of the art, but it won't reverse the flow in that line at 25 psig off-system delivery pressure either.
The two big names I run into at a lot of my clients are PipeFlow and Gap (usually linked with the wellbore simulator "Prosper" and the reservoir simulator "MBAL"). I've found PipeFlow to crash in very unpleasant ways with complex networks (e.g., much above 200 nodes and it can crash in a way the corrupts your input file and you either have to fall back to a backup or start over, I create a new backup before every time I hit "calculate"). Gap is pretty expensive and I've had a really hard time calibrating models in it. It seems that small changes make a big difference, but big changes make no difference. It can be frustrating.
If the lowest pressure you anticipate is above probably 100 psig, then I've had pretty good results from Schlumberger's PIPESIM. It is pretty intuitive and most of the obvious things that limit other programs are not problems with it. If it has a major bust, I haven't found it yet. For pressures less than that I'm still looking for something that will replace my MNET.
David Simpson, PE
MuleShoe Engineering
"Belief" is the acceptance of an hypotheses in the absence of data.
"Prejudice" is having an opinion not supported by the preponderance of the data.
"Knowledge" is only found through the accumulation and analysis of data.
The plural of anecdote is not "data"
